1 /**********************************************************************
2 iso8859_9.c - Oniguruma (regular expression library)
3 **********************************************************************/
4 /*-
5 * Copyright (c) 2002-2007 K.Kosako <sndgk393 AT ybb DOT ne DOT jp>
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 */
29
30 #include "regenc.h"
31
32 #define ENC_ISO_8859_9_TO_LOWER_CASE(c) EncISO_8859_9_ToLowerCaseTable[c]
33 #define ENC_IS_ISO_8859_9_CTYPE(code,ctype) \
34 ((EncISO_8859_9_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)
35
36 static const UChar EncISO_8859_9_ToLowerCaseTable[256] = {
37 '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007',
38 '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017',
39 '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027',
40 '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037',
41 '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047',
42 '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057',
43 '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067',
44 '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077',
45 '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
46 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
47 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
48 '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137',
49 '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147',
50 '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157',
51 '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167',
52 '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177',
53 '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207',
54 '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217',
55 '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227',
56 '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237',
57 '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247',
58 '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257',
59 '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267',
60 '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277',
61 '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
62 '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
63 '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\327',
64 '\370', '\371', '\372', '\373', '\374', '\335', '\376', '\337',
65 '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347',
66 '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357',
67 '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367',
68 '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377'
69 };
70
71 static const unsigned short EncISO_8859_9_CtypeTable[256] = {
72 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
73 0x4008, 0x420c, 0x4209, 0x4208, 0x4208, 0x4208, 0x4008, 0x4008,
74 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
75 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,
76 0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
77 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
78 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,
79 0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,
80 0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,
81 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
82 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,
83 0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,
84 0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,
85 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
86 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,
87 0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,
88 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
89 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
90 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
91 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,
92 0x0284, 0x01a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,
93 0x00a0, 0x00a0, 0x30e2, 0x01a0, 0x00a0, 0x01a0, 0x00a0, 0x00a0,
94 0x00a0, 0x00a0, 0x10a0, 0x10a0, 0x00a0, 0x30e2, 0x00a0, 0x01a0,
95 0x00a0, 0x10a0, 0x30e2, 0x01a0, 0x10a0, 0x10a0, 0x10a0, 0x01a0,
96 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
97 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,
98 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,
99 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,
100 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
101 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,
102 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,
103 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2
104 };
105
106 static int
mbc_case_fold(OnigCaseFoldType flag,const UChar ** pp,const UChar * end ARG_UNUSED,UChar * lower)107 mbc_case_fold(OnigCaseFoldType flag,
108 const UChar** pp, const UChar* end ARG_UNUSED, UChar* lower)
109 {
110 const UChar* p = *pp;
111
112 if (*p == 0xdf && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
113 *lower++ = 's';
114 *lower = 's';
115 (*pp)++;
116 return 2;
117 }
118
119 *lower = ENC_ISO_8859_9_TO_LOWER_CASE(*p);
120 (*pp)++;
121 return 1;
122 }
123
124 #if 0
125 static int
126 is_mbc_ambiguous(OnigCaseFoldType flag, const UChar** pp, const UChar* end)
127 {
128 int v;
129 const UChar* p = *pp;
130
131 if (*p == 0xdf && (flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {
132 (*pp)++;
133 return TRUE;
134 }
135
136 (*pp)++;
137 v = (EncISO_8859_9_CtypeTable[*p] & (BIT_CTYPE_UPPER | BIT_CTYPE_LOWER));
138 if ((v | BIT_CTYPE_LOWER) != 0) {
139 /* 0xdf etc.. are lower case letter, but can't convert. */
140 if (*p >= 0xaa && *p <= 0xba)
141 return FALSE;
142 else
143 return TRUE;
144 }
145
146 return (v != 0 ? TRUE : FALSE);
147 }
148 #endif
149
150 static int
is_code_ctype(OnigCodePoint code,unsigned int ctype)151 is_code_ctype(OnigCodePoint code, unsigned int ctype)
152 {
153 if (code < 256)
154 return ENC_IS_ISO_8859_9_CTYPE(code, ctype);
155 else
156 return FALSE;
157 }
158
159 static const OnigPairCaseFoldCodes CaseFoldMap[] = {
160 { 0xc0, 0xe0 },
161 { 0xc1, 0xe1 },
162 { 0xc2, 0xe2 },
163 { 0xc3, 0xe3 },
164 { 0xc4, 0xe4 },
165 { 0xc5, 0xe5 },
166 { 0xc6, 0xe6 },
167 { 0xc7, 0xe7 },
168 { 0xc8, 0xe8 },
169 { 0xc9, 0xe9 },
170 { 0xca, 0xea },
171 { 0xcb, 0xeb },
172 { 0xcc, 0xec },
173 { 0xcd, 0xed },
174 { 0xce, 0xee },
175 { 0xcf, 0xef },
176
177 { 0xd0, 0xf0 },
178 { 0xd1, 0xf1 },
179 { 0xd2, 0xf2 },
180 { 0xd3, 0xf3 },
181 { 0xd4, 0xf4 },
182 { 0xd5, 0xf5 },
183 { 0xd6, 0xf6 },
184 { 0xd8, 0xf8 },
185 { 0xd9, 0xf9 },
186 { 0xda, 0xfa },
187 { 0xdb, 0xfb },
188 { 0xdc, 0xfc },
189 { 0xdd, 0xfd },
190 { 0xde, 0xfe }
191 };
192
193 static int
apply_all_case_fold(OnigCaseFoldType flag,OnigApplyAllCaseFoldFunc f,void * arg)194 apply_all_case_fold(OnigCaseFoldType flag,
195 OnigApplyAllCaseFoldFunc f, void* arg)
196 {
197 return onigenc_apply_all_case_fold_with_map(
198 sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
199 flag, f, arg);
200 }
201
202 static int
get_case_fold_codes_by_str(OnigCaseFoldType flag,const OnigUChar * p,const OnigUChar * end,OnigCaseFoldCodeItem items[])203 get_case_fold_codes_by_str(OnigCaseFoldType flag,
204 const OnigUChar* p, const OnigUChar* end, OnigCaseFoldCodeItem items[])
205 {
206 return onigenc_get_case_fold_codes_by_str_with_map(
207 sizeof(CaseFoldMap)/sizeof(OnigPairCaseFoldCodes), CaseFoldMap, 1,
208 flag, p, end, items);
209 }
210
211 OnigEncodingType OnigEncodingISO_8859_9 = {
212 onigenc_single_byte_mbc_enc_len,
213 "ISO-8859-9", /* name */
214 1, /* max enc length */
215 1, /* min enc length */
216 onigenc_is_mbc_newline_0x0a,
217 onigenc_single_byte_mbc_to_code,
218 onigenc_single_byte_code_to_mbclen,
219 onigenc_single_byte_code_to_mbc,
220 mbc_case_fold,
221 apply_all_case_fold,
222 get_case_fold_codes_by_str,
223 onigenc_minimum_property_name_to_ctype,
224 is_code_ctype,
225 onigenc_not_support_get_ctype_code_range,
226 onigenc_single_byte_left_adjust_char_head,
227 onigenc_always_true_is_allowed_reverse_match
228 };
229
